When configuring memory for a VM, make GB be the default instead of MB. These days I think everyone would save some clicks and time if this was implemented.

Cuando se activa la memoria dinamica en las VMs, se afecta el UI de la memoria RAM sobre el "Administrador de tareas" y los contadores de rendimiento.
El tope del porcentaje de la memoria, debería seguir siendo el maximo y no el asignado por el VMHost, Ademas de esto, la tasa de consumo no debería verse afectado por el consumo del controlador, esto debería ser ignorado ya que, cuando se analiza el consumo de una VM dentro de ella, no se sabe realmente cuanto esta consumiendo sin programas adicionales como RAMMAP.

Currently when you create a VM in VMM it adds a prefix and suffix to the role name when created in Failover Cluster Manager. "SCVMM ServerName Resources" for example.

When I name a VM, I want the name to stay the same no matter where I am looking. (Hyper-V Manager, Failover Cluster Manager and VMM.) If I name a server "JoesWebServer" I don't want it to change in FCM to "SCVMM JoesWebServer Resources".

The Logical Switch name are limited to 64 characters.
For the implementation with CISCO APIC we need more as 64 characters e.g. 128 characters for the Logical Switch name.
The CISCO APIC delpoyed the Logical Switch with a compex naming convention e.g. TenantName | APIC Name | EPG Name | Logical Switch Name.

Configuring the network adapter of the bare-metal is configurable only after deep discovery has been made. like makeing the nic not registering with DNS option.

Physical Computer Profile mostly does alot of time, but only able to preconfigure the network adapter in detail.

When you go apply a classification to a VM's Virtual Disk in the GUI, the drop-down menu sorts things in an incoherent manner.

It appears that Get-SCStorageClassification also returns results in the same order.

It also does not appear to sort based on propreties ID, AddedTime, or ModifiedTime.

Can we at least get the VMM GUI to sort results? (Powershell can obviously be done manually.)

Image attached demonstrating issue.

Customer is sometimes 'ok' state. And other times, 'overcommit' state, even without changes in environment.

SCVMM 2016 UR4, Standalone VMM environment of 12 Nodes.
He has 192 GB free when he calculates all his HOST's free memory including the reserve node's (R1).

https://support.microsoft.com/en-us/help/3023928/details-of-the-cluster-overcommit-algorithm-in-system-center-2012-r2-v

• What we might really need here is logging of how 'ok' and 'overcommit' was calculated so the toggle between both states is understood.

Currently VMM is using unattend.xml for domain joins for VMs and Hyper-V hosts. This contains having unencrypted passwords stored as cleartext at some points of time during the deployment.

We would prefer using offline domain join instead - as it was done for nano server in the past.

Add VHD Set support in VMM. Support for shared VHDX and now VHD Set has been requested for years, yet VMM ignores this critical feature.

One of the greatest advancements of VHD Set over shared VHDX is the ability to migrate vhd set, yet VMM has no support for this.

Please add VHD Set management (create, attach, dettach, migrate ....) to VMM.

When deploying a VM from a template, I rename the boot disk the template is made from to "system.<fullvmname>.vhdx" because I do not want the disk created as "<vmtemplatename>.vhdx. If there was a way to setup a rule to do this, it would be nice.

Since Windows Server Container Support. It should be possible to create Containers and Hyper-V Containers with SCVMM. Also suggesting to integrate Kubernetes Management Board Integration into SCVMM since Kubernetes Support Windows Containers.

When pushing a VMM 2016 UR4 agent to an infrastructure server. If the infrastructure server has an existing certificate bound to HTTPS, example RDSGateway and Network Controller. The agent installation will remove the existing certificate binding when setting up BITS.

It would be great if the Operating System list was something the administrator could define, but in lieu of that, add FreeBSD and OpenBSD as additional OS options for VMs.

I find it frustrating when I'm moving a lot of resources around, to have to clear out the "missing" items from the library. Library should be just treated like file explorer essentially. When its gone, its gone. If people like a audit trail, the offer a "keep history for x days" then show as missing......for x days. That would be better.

The requirement for running sysprep when preparing VM Templates (with Windows) looks a bit confusing, because as soon as you start converting finalized VM, the status changes to 'Sysprepping...'.

If VMM could do sysprepping on it's own, I'd like to be able to skip doing this step in my Task Sequence, because one of the last steps I'm doing is applying some hardening GPOs, which in turn, cause sysprep to succeed only after a subsequent reboot (and automating one is quite a challenge as apparently MDT doesn't like something that's being touched by the hardening policies)

I have different libraries in different locations, on different speed networks. Some locations are deduplicated, some are not.

For me, if I want to move an item from the library to another, I have to "deploy" to a cluster, then "store in library" again to the destination library.